AA Intelligence Index v4.3. Source median output speed. Test configuration is not reported by the Free API; throughput is not response latency. Source captured 2026-10-18T06:15:17.223Z. Historical values are preserved separately.

Qwen3 8B (Non-reasoning)

Provider: Alibaba

General intelligence index

6

Output throughput (tokens/second)

Unavailable

Input USD per million tokens

0.2

Output USD per million tokens

0.2

Combined context tokens

32768

Context evidence

Publisher model card · View source

Maximum input tokens

Unavailable

Maximum output tokens

Unavailable

Conditional extended context

131072 tokens. Native context from the publisher's model card. Extended context requires the documented RoPE/YaRN configuration. Recommended generation lengths are not treated as hard API output limits.

Output throughput is not time to first token. A listed zero rate is not proof of a free offer.

Data provided by artificialanalysis.ai. Context limits: the source attached to each model. General model measurements do not establish healthcare or legal suitability.